The Carbon Footprint Impact of Cloud Solutions
Contributed Commentary by Mladen Lazarevic, Seven Bridges
February 3, 2023 | Researchers turn to bioinformatics computing to analyze large datasets and tackle issues ranging from precision medicine solutions in oncology to novel drug therapies. This is especially true of genomic data and bioinformatics analysis.
While those in the life sciences appreciate the importance of curbing greenhouse gas emissions and the pressing impact of global warming, actionable steps toward going green have yet to permeate all facets of cloud computing research. As such, the current state of affairs in bioinformatic computing is not particularly green.
The scale of data sets analyzed in bioinformatics calls for large processing centers to analyze the data. Historically, there was no push to be green in research, so efficiency and accuracy—not minimizing carbon emissions—were prioritized. However, a recent study quantified the scope of the carbon footprint in bioinformatics, alluding to the benefits of cloud computing options for bioinformatic analysis (bioRxiv, DOI: 10.1101/2021.03.08.434372).
Though most bioinformatics processing centers are not carbon net zero, some have made strides in that direction. For example, Google, Amazon, and Microsoft have initiatives to become greener in their computing endeavors. In some cases, companies will offset carbon emissions by planting trees or using other methods. However, most are not yet carbon net zero but rather on a trajectory toward carbon neutrality.
Go Green to Make Green
Several factors drive this push towards carbon-neutral bioinformatics analysis options like cloud computing. For one, there’s the philanthropic side. Companies must do their part to curb emissions. The logical factor of addressing global warming from a pragmatic standpoint—we need an inhabitable planet to conduct bioinformatic analysis—also plays a part. However, a third and not as commonly considered factor really should drive this push: economics. Companies are for-profit businesses, and greener bioinformatic analysis options can benefit companies in terms of their bottom line.
A significant benefit of going green is gaining green: companies can save money in the long run by turning to more carbon-friendly options for their bioinformatic analysis. For example, turning to cloud computing for bioinformatics can decrease costs since companies do not need to analyze in-house but can outsource to a third party to maintain the data repository and cloud computing infrastructure. In addition, having cloud solutions for bioinformatics can make the data more easily accessible to collaborators to run analyses. Of course, keeping the data in the cloud source is still a cost, but it is ultimately a more economical option for companies.
Then there’s the increased efficiency afforded by cloud computing to consider. It has been noted for years that some data centers waste 90% of the power they pull from a grid by running systems at full speed, regardless of demand. Cloud computing options thus offer a more effective solution as they run analysis on demand, not as a matter of course.
But it’s not just the financial soundness of cloud computing that makes it worthwhile to companies: cloud computing for bioinformatics can deliver improved results when paired with a third-party vendor for bioinformatic analysis. With cloud computing, new tooling is updated regularly by the third party, ensuring the analysis is run on the most up-to-date, precise technology. Ideally, each iteration of the tooling improves results. Otherwise, it would not be implemented; therefore, running data on improved versions means better analysis.
In addition, cloud computing allows for collaboration by sidestepping some of the silos typically found in research endeavors. For example, researchers on various projects can all use the same data via the cloud rather than hunting down bioinformatics data stored with just one department or another. Ultimately, this equates to shortening the research timeline, which means precision medicine solutions get to those who need them faster.
Speed Bumps in the Way of Green Computing
Despite the clear benefits of moving to greener, more carbon-neutral cloud computing options, several factors stand in the way of—or at least are slowing down—companies from leaping. But they can all be boiled down to some variation of this: growing pains.
Companies need to invest time, manpower, and patience in dealing with the frustration of moving data from a center to the cloud. That means navigating new technology and additional effort when the current method is less green. That can be an uphill battle and potentially not worth fighting if a business only looks at the short term. However, on a long enough (and not very long at that) timeline, the move pays off in spades.
If opting for a tool update, another hurdle to note and is worth considering is this: data reproducibility. Though this is not a common problem, it is worth reviewing. Experiments and analysis run on past iterations of tooling may yield different results, though the upgraded tech may be more accurate and more carbon friendly. In rare instances, this can come to a head regarding U.S. Food and Drug Administration approvals and compliance issues that may arise. For example, a company may need to resubmit for approval if methods change beyond the scope of clearance initially granted. That could prove a headache in terms of time and money to go through the approval process again. Of course, companies can use the original tooling in the cloud to circumvent this potential issue.
The Future is Green and in the Cloud(s)
Despite the challenges, the field is trending toward greener options for bioinformatics cloud computing. Partners have noticed this trend, and some already offer systems to ease the growing pains of going green and cloud based. For example, some companies market turnkey support to help businesses move their data to cloud computing options, especially in bioinformatics, where the datasets are substantial.
In addition to resources available to support companies in making the plunge, there are immediate benefits of making the shift that draw businesses to go green. For example, having another company be responsible for updating the tooling eliminates the burden, lessening some of those growing pains. This allows companies to focus their energy on research, not maintenance. And, of course, in the end, many companies have their eye on the prize: getting bioinformatic analysis done more quickly and effectively so treatments and therapies can get to the market and help people. After all, the end game is driving precision medicine solutions.
Going green and cloud-based is another advantage that offsets those growing pains: democratizing access to data. Companies can more effectively collaborate across the organization by eliminating silos via cloud computing options for their bioinformatics research. As a result, they should enjoy more equitable sharing of data and potentially shortened timelines.
But it all comes down to this: We only have one planet, and computing is a necessary aspect of our work here, so we need to do our part in bioinformatics to preserve it.
Mladen Lazarevic is the associate director of bioinformatics at Seven Bridges, with several years of experience in bioinformatics data analysis, team and department management, programming, data science methods, and signal processing. In addition, he has experience in multi-omics areas, cancer bioinformatics, immunology, SNV calling, genome assembly, and developing and implementing NGS pipelines in an optimized and scalable manner on Amazon Web Services and Google infrastructure. He can be reached at mladen.lazarevic@sbgenomics.com.